Estimating acacia mangium plantations standing timber volume using an unmanned aerial vehicle

Abstract

Forest plantations are anticipated to play the important roles to maintain the commercial supply of logs and reducing timber demand from the natural forest. Thus the estimation of timber volume is crucial for the management to assess their productivity potential. Generally, the estimation of timber volume is considered a difficult task as ground measurement is expensive. The extensive use of Unmanned Aerial Vehicles (UAV) in forest plantations and indirect measurements of basic tree attributes is suitable for monitoring and timber volume estimation. The UAV measurement has the advantage that it can acquire the high resolution digital imagery through photogrammetry and Structure from Motion (SfM) to estimate basic tree attributes such as tree heights, diameter at breast height (DBH) and crown diameters. The purpose of this study is to estimate the timber volume of Acacia mangium plantations using ortho-mosaic, digital surface model (DSM) and digital terrain model (DTM), which are derived from UAV data. The involved workflow used Canopy Height Models (CHM) for the tree height extraction, the smoothing of raster images to find the local maxima and Inverse Watershed Segmentation (IWS) for estimation of the crown diameters which indirectly used to estimate the DBH and the calculation of Acacia mangium timber volume. All this workflow is done with the help of geographical information system. Finally, the accuracies of the two methods are validated by testing their significance. Results showed higher agreement between field measured and estimated UAV data for tree heights than for DBH based on RMSE, which ranged from 1.97 m – 3.33 m for height and 3.67 cm – 4.49 cm for DBH. However, the Pearson’s r for DBH derived from UAV measurement data shows weak correlation than that of the field measurement which ranged from 10% – 48%, while tree height shows the strong correlation between UAV measurement and field measurement which ranged from 75% – 99%. For the timber volume estimation, the RMSE ranged from 0.25 m3 – 0.34 m3 and Pearson’s r ranged from 33% - 62%. In overall, the accuracy of the results is acceptable and showed that the methods were feasible for Acacia mangium timber volume estimation.